论文部分内容阅读
文章采用时间序列分解模型对我国2012年至2016年各个季度的空调销售量进行预测。先确定并分离季节成分;然后对消除了季节成分的序列选择三次曲线回归预测模型,并利用该模型对销售量进行预测;最后利用求得的预测值乘以相应的季节指数,得到含有季节性因素的最终预测值。在此基础上对我国2017年各季度的空调销售量进行了预测。
The article uses the time series decomposition model to predict the sales of air conditioners in all quarters from 2012 to 2016 in our country. First determine and separate the seasonal components; then choose the cubic regression regression model to eliminate the seasonal components and use the model to predict the sales volume; finally, multiply the corresponding seasonal index by the obtained forecast value to obtain the seasonal The final predictive value of the factor. Based on this, we forecast the sales volume of air conditioners in each quarter of 2017 in our country.